This study began as an efficacy study of IFN-2alpha in pts. with hairy cell leukemia. It was observed that most pts. responded to IFN, but that very few complete responses were being obtained. Studies being done elsewhere confirmed the low complete remission rate. Once IFN was stopped, nearly uniformly disease progression requiring reinstitution of therapy was observed. There appear to be very few if any pts. who will not require treatment. Because of these findings, we opted to administer IFN continuously to pts. who were initially responsive to this drug. Of the 53 evaluable pts. (of the 56 entered on this study), there was 1 complete remission, 41 partial remissions, 1 minor response, 9 pts. with stable disease and only 1 pt. with disease progression. 14 pts. continue to receive IFN without interruption with a median duration of continuous IFN treatment of 9.2 years. 34 pts. discontinued IFN for a variety of reasons, the most common being the development of acquired IFN resistance in association with IFN antibodies. The resistance to IFN was manifested early, in the first 18 months of treatment, except in 2 cases. An important finding in this study is the continued slow, but significant, hematologic improvement in absolute granulocyte and platelet counts beyond 18 months of therapy, thereby indicating that prolonged treatment results in continued benefit rather than the production of antibodies with subsequent development of IFN resistance. Although it is clear from this study that hairy cell leukemia can be controlled in the long-term with IFN, longer follow-up will be necessary to determine if continuous therapy with IFN is better than intermittent therapy.